January, 2009: I wrote this essay about 10 years ago. I believe what is written below regarding steel, aluminum and titanium remains true. Carbon, however has become the preferred material for the high performance bike because it is now so light, so strong and so very, very reliable. Yet, I remain convinced that the ride quality of carbon is still greatly inferior to that of steel.
Bicycle Frame Materials:
Steel, Aluminum, Ti or Carbon?
Here is the basic thesis of this essay: there are three forces that have recently and profoundly changed the design of the road-racing bicycle
- The usual search for competitive advantage on the race course.
- Marketing pressures.
- The manufacturer's search to reduce costs, and the corollary desire to reduce the number of items manufactured.
The result? Most people are being sold bicycles that have had their ride quality seriously compromised in the name of increased performance that is really illusory. To put it bluntly, the bikes they are being sold ride like bloody hell. The pleasure that can be had by riding a delicious, well-designed, comfortable bike that is competitive at all but the very highest levels of racing has been denied to the majority of the modern road bike riders.
A discerning rider should look upon lightweight road bikes as basically falling into two categories.
- All-out racing bikes (and lower-priced clones that purport to be the same, but have none of the competitive advantages and all of the disadvantages of these bikes) that, while not offering the most comfortable or pleasurable ride, give the serious athlete a chance to race without worrying that his equipment will hold him back. Many items that claim to be in this category are marketing tools aimed at the misinformed. But more about this later.
- Riding bikes. An intelligently designed bike will offer a wonderful ride, handle like a dream and only slightly compromise the all-out need for the ultimately competitive bike. I believe that these are the bikes most people should ride.
This all used to be so easy.
In 1975, a professional rider rode a bike made of Columbus, or Reynolds 531. The usual group was Campagnolo Nuovo Record, but a mixture of French or even Spanish components could be used. The rims would be tubulars like Fiamme strung up with 36 spokes, and have handmade (cold-treated) tubulars glued on. The bars were most likely Cinelli, the chain and cogset Regina. The bike probably weighed about 21.5 pounds. It was the state of the art. No bike could be made that was faster, more reliable or significantly lighter.
At the same time, no bike could be had that rode better than a handmade frame, built by a master out of 531 or Columbus and assembled with the pro equipment of the day. The bike was not only competitive at the highest levels of the sport, it rode comfortably and had excellent vertical compliance so that it adhered to the road. The sensual element that gives a bike ride its real pleasure could find no better tool than the true pro bike of the '70s.
Bikes have changed. Those changes require that a thoughtful person contemplating the purchase of a new bike to use a different set of criteria than a rider of the '70s and '80s. It's no longer enough to just want what the pros ride.
As I stated in my opening paragraph, there are several forces at work changing the design of the professional racing bike. First of all, the most easily noticed is the march to lighter weight. This has caused the UCI (the governing body of worldwide bike racing) to require that bikes weigh at least 6.8 kilograms (15 lbs.). There is a fear that the technological advance will obscure the importance of the human effort in a bike race. The UCI wants to make sure that the emphasis is on the athlete.
There are two other forces at work. Manufacturers want to reduce the number of different items that they manufacture. Sometimes this ends up being for the better. Threadless forks are lighter, the stems are stiffer. The manufacturer gets to make only one fork and cut it to length. Previously he had to make a different threaded fork for each size frame. His life is simpler and the bike is lighter and stiffer.
Compact frames reduce the number of frame sizes that need to be manufactured. We make the Corsa Strada (not a compact frame) in 18 sizes, each with a unique geometry. This many items, all complex and differernt, are a manufacturer's nightmare, but a bike rider's dream.
Fit and weight distribution on compact frames, however, can be a problem. If a person is between frame sizes, the choice between a too-short or too-long stem (and the same with the seat post) will result in a poor fit and bad handling. The miniscule weight savings does not make up for the disadvantages. But, the manufacturer has saved a bundle. And you'll look like the pros in the Tour de France who are being paid to sell this line of baloney.
Another force driving design is advertising. Marketing costs are huge. In the sport-driven world of bicycle advertising, these costs have ballooned. Up until the 1970s, a big bicycle factory could, on its own, sponsor a high-end bicycle team. Gone are the days of a Peugeot team stocked with Tour winning riders like Thevenet. The last team sponsored primarily by a bicycle company was TI-Raleigh. And this was possible because Tubing Investments, the owner of Raleigh, amortised a lot of the costs of the team over other branches of the conglomerate. Today, the bicycle supplier generally plays a much smaller part in the financing of a pro team, which can cost more than $8,000,000 a year. With these huge costs, the pro team has been forced to become a more focused marketing driven device than in decades past. Every square inch of the bike and the riders' clothing is crammed with advertising.
In the past, rim manufacturers never paid racing teams money to equip their bikes because the then-used box-section rim is so anonymous. You never knew which teams were using what rims. New wheels with deep section rims can carry obvious and easily recognized advertising, but this is at a real cost to the rider. The deep rims ride very harshly. The reduced spoke count wheel have very high spoke tensions that exacerbate the problem.
The consumer is sold the wheels because they are light. But, this is a half-truth. Because the rims have deep sections, the inertial mass (rotating weight) is greater. The result is that the bike has less snap and rides more harshly. To make it worse, tests have shown that a rim needs to be 40 mm deep to have any real aerodynamic advantage. The rims with their cross sections in the 30mm's are not aero, they are only fancy looking. Before buying one of these wheelsets that have lots of gee-whiz, consider a nice set of 32 hole, cross-three wheels with box section rims. Put a pair on your bike and give them a chance. Borrow a buddy's set if you have any. The weight is almost the same. But the ride.........
On to the frames.
Bicycle frames and forks are made of four different materials: steel, aluminum, titanium, and carbon fiber. There other materials on the fringe, but they are beyond the scope of this discussion.
Steel, aluminum and titanium are metals with an interesting relationship. Titanium has 1/2 the density of steel and 1/2 the tensile strength. Aluminum has 1/3 the density and 1/3 the strength of steel. Now the obvious conclusion that one can draw from this is that a frame of aluminum should end up weighing the same as a steel frame. A given cubic volume of aluminum weighs 1/3 as much as the same volume of steel. But because it has 1/3 the strength, one would need 3 times the material to make a given structure work.
It's not that simple. In building bicycles, made of 9 tubes, there are problems in drawing the tubes too thin in relation to their diameter. Given the current state of the art, a bicycle downtube of steel is roughly 32 mm in diameter. It can be drawn so that the walls are about 0.4mm thick. Any thinner and the tube can buckle easily, and the tube is subject to denting as well.
Modern aluminum downtubes are usually oval for greater resistence to bending under the normal load of pedaling. The mechanical characteristics of aluminum are inferior to steel. It will fail more quickly under repeated stress. So, the diameters and wall thicknesses must be increased, but not as much as would be called for to make up entirely for the reduced tensile strength of aluminum. With a 42mm diameter, aluminum tubes can be drawn down to 0.7mm yielding a tube that weighs 185 grams, compared to 220 grams for a state-of-the-art steel downtube. Multiplied throughout the bike's 9 tubes, it is obvious that a significant weight saving can be achieved using non-ferrous (non-steel) materials. Intermediate weight savings can be gained using titanium.
We know more about steel than any other of the materials used to build bikes. It is one of the basic building blocks of our civilization. Even though it has been over 150 years since Henry Bessemer figured out how to produce steel commercially and cheaply, new developments keep coming. In the early 1990s, Columbus announced the introduction of Nivachrome steel. Previous steels used to build bikes were chrome-moly alloys that lost as much as 40% of their strength after brazing. Nivachrome was the first alloy specifically developed for building bicycles. In the hands of a competent builder, Nivachrome loses only 10% of its strength after building. Also, it is a highly ductile steel (not brittle like glass) and has a very high tensile strength. The results? Steel tubes could be made thinner and lighter.
Previously, because so much strength was lost in brazing or welding, the tubing maker had to put a lot of redundant material in the tubing to allow for the loss of strength. A normal tube in 1976 was 0.9mm at the butted end and 0.6 in the center. With Nivachrome, that changed to 0.7mm at the butt and 0.4 in the center. This is a reduction of 1/3 of the mass of the tube with no loss of strength or durability.
The resulting bikes made with these modern 0.4mm thick tubes were light and had a new, beautiful, elastic, delicious sensual feel that I cannot describe. I can only tell you that it is there.
The other advantage of steel is reliability. Modern metallurgy teaches that if a steel tube is bent and released repeatedly, it will not break as long as the amount of this bending remains within what engineers call the "elastic" range of the metal. When a steel tube is bent until it permanently deforms, it is said to have failed. As long as the frame is never bent beyond it's elastic range, its life will be very long, indeed.
But let us not deceive ourselves. The slight weight disadvantage that comes with a steel frame makes it unusable for racing at the highest levels. A steel frame can be made that weighs in the mid to low 3-pound range. Over a non-compact aluminum frame, this is a penalty of about one pound. This is just too heavy to chase Tyler Hamilton up a category-one climb. That is why the professional peloton uses aluminum or carbon. But for the rider who does not compete at the elite level, that one-pound penalty as part of a whole rider/bike package that approaches 200 pounds (or may generously exceed it) is insignificant. And for that pound, the rider gets a bike that can take advantage of the high-tensile strength and springy elasticity of modern steel and ride a bike that is an absolute dream. No bike rides as well as a steel bike built by a skilled builder. People who disagree with this conclusion usually have either a commercial interest in other materials, or have not ridden modern steel bikes.
Forks. Putting on a full carbon fork will save almost a pound. This is a huge weight savings in a single component for a very modest price. I will save most of my comments for the carbon discussion below. But, because the deformation of steel under load is linear, the springy feel of a steel frame is greatly enhanced by the lively feel of a steel fork.
The evolution of the aluminum frame has been going at the speed of light. The first modern aluminum frames that showed up in the mid 1970s by Alan of Italy used the same diameters that were then used in making steel frames. The resulting frames were very light, but they were so soft, I considered them unrideable. Vitus also made aluminum frames that were very soft and very light. I could say these were terrible bikes, but Sean Kelly beat the rest of the world like gongs with these spongy imitations of real bikes. There are many roads to the same place. One man's food is another man's poison.
Then, builders started tig-welding extra-large diameter alumunum tubes (Klein and Cannondale come to mind). The resulting bikes were fantastically stiff, very light, but hard as nails. I considered them terrible alternatives to good steel bikes because they had gone too far the other way. The ride was just too harsh.
The engineers never stopped working on solving the problems aluminum presented. Columbus came out with Altec and the ride got better. Easton's offerings made frames that I finally admitted were tolerable. Columbus' modern tubes Altec2 and Airplane make fine riding bikes. I don't put their ride quality in the class of the good steel bikes. But, if a rider is seeking competitive light weight and acceptable ride quality, a non-compact, horizontal top-tube aluminum bike will serve him well. The earlier problems of aluminum's tendency to fail after only a short time has been basically solved. But the lifetime of an aluminum frame is not and will not be that of a steel bike. The rider has to accept that in his search for high-perfomance, compromises must be made. These are not lifetime bikes. They just aren't!
This is the middle ground. Titanium is not as strong as steel and it is not as light as aluminum. The result is a frame that has a better ride than aluminum, but weighs more. It doesn't ride as well as steel, but it is lighter. It is very reliable. Titanium frame failures, like steel frame failures, are very rare. Titanium also has another advantage. It doesn't rust. Riders in areas where roads are salted like titanium frames. They don't have to be painted. If the decals get wrecked, new ones are easy to apply.
I have good friends whose judgement I respect who insist that this is the best material from which to build a frame because of its middle ground between aluminum and steel. At this point, it becomes a matter of taste.
I think it was about a decade ago that Greg LeMond predicted that carbon fiber bikes were the future. I don't think they have yet lived up to that promise, even with Lance Armstrong's four (at the time of this writing) Tour de France wins on carbon.
The first carbon frames I recall were from Alan and Vitus. The aluminum tubes that were usually glued into aluminum lugs were replaced with carbon tubes. Then, the next big advance came from Kestrel with a monocoque frame. Lots of riders went wild over them and adored the ride and lightness. Like all early products, all of the above frames suffered from growing pains. Reliability just wasn't there.
If we fast-forward to today, the advances have been profound. The basic problem of carbon's reliabilty has been solved. Millions of men and women trust carbon for the their forks, the component on a bike that requires the greatest level of quality and reliabitlity. A broken fork can be catastrophic.
Producers of carbon fiber products have advanced the basic technology of the raw material. Carbon fiber products are made of carbon fiber imbedded in an epoxy resin. As the technology has advanced, engineers have been able to reduce the amount of resin as well as cabon fiber in the basic make-up of a carbon item.
This is important not only for a weight savings, the basic reason people turn to carbon, but also for road feel.
I have a theory.
When the first carbon tubes frames from Alan and Vitus showed up, they were slightly over size. They were still quite soft as frames go. Then, when the first carbon fiber forks were introduced, many of them mimicked the dimensions of steel forks. They were terribly underdesgined. They were so soft that the front wheel would vibrate back and forth in an unnerving shimmy when the front brake was applied. These very soft, improperly designed items gave carbon a reputation for softness. If an item were carbon, it was assumed that it would give a soft, comfortable ride. It was not the carbon, which is an extremely stiff material, that was causing the softness, it was the engineering.
Look at carbon forks today. They are built to dimensions that are appropriate to the material. Carbon forks are stiff, reliable and well-made.
So how does carbon fiber affect the ride of the bike? Engineers can do almost anything they want with carbon. It's all in how the sheets of woven carbon fiber are laid and epoxied together. There is, however, a great deal of commonality in how all carbon fiber products are designed for bicycles. So, we can paint with a broad brush and made some general rules.
The epoxy in the carbon fiber matrix damps high-frequency vibrations. If you have a nice road, it will take out that road buzz and smooth it out.
However, because the deformation of carbon under load is not linear, the worse the jolt or impact, the harsher the feel. It can make a good road nice, but it will make a bad road terrible as it bites back when the usual elastic limit of the carbon is reached.
Some riders find the deadening of road buzz delightful and pleasant. Other find that carbon fiber results in a dead feel. Once again, it's a matter of taste.
I believe that the feel of the road is a large part of the feedback I am looking for as I ride my bike. I am not looking to isolate myself from my cycling environment. I want to be part of it. For me, then, carbon works against my cycling goals. I have never ridden a carbon frame or fork that gives the fine, pleasant, comfortable ride under the widest set of condition that steel gives.
With changing technology, this judgment is only for today.
So, what kind of bike to buy?
I still hold that a modern steel bike built with hand-made cross-three wheels and good tires inflated to a maximum of 110 psi remains the gold standard. No other bike will ride as well.
The elite racer looking for the lightest, fastest bike is best served with a lightweight carbon bike. If he avoids compact frames (now almost impossible), he will get a finer ride at almost no weight gain. The important, and now ignored, questions of fit and weight distribution are better answered with a horizontal-top-tubed bike.